Reduced-Dimensionality Semiclassical Transition State Theory: Application to Hydrogen Atom Abstraction and Exchange Reactions of Hydrocarbons.

Subscriber access provided by UNIV OF MISSISSIPPI

Article

Reduced-Dimensionality Semiclassical Transition State Theory: Application to Hydrogen Atom Abstraction and Exchange Reactions of Hydrocarbons Samuel Greene, Xiao Shan, and David Clary J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.5b04379 • Publication Date (Web): 19 Jun 2015 Downloaded from http://pubs.acs.org on June 23, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Reduced-Dimensionality Semiclassical Transition State Theory: Application to Hydrogen Atom Abstraction and Exchange Reactions of Hydrocarbons Samuel M. Greene, Xiao Shan, and David C. Clary* Physical and Theoretical Chemistry Laboratory, Department of Chemistry, University of Oxford, South Parks Road, Oxford, OX1 3QZ, United Kingdom *E-mail: [email protected]; Telephone: +44 01865 276100

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 47

Abstract Quantum mechanical methods for calculating rate constants are often intractable for reactions involving many atoms. Semiclassical Transition State Theory (SCTST) offers computational advantages over these methods but nonetheless scales exponentially with the number of degrees of freedom (DOFs) of the system. Here we present a method with more favorable scaling, Reduced-Dimensionality SCTST (RD SCTST), that treats only a subset of DOFs of the system explicitly. We apply it to three H abstraction and exchange reactions for which two-dimensional potential energy surfaces (PESs) have previously been constructed and evaluated using RD quantum scattering calculations. We differentiated these PESs to calculate harmonic frequencies and anharmonic constants, which were then used to calculate cumulative reaction probabilities and rate constants by RD SCTST. This method yielded rate constants in good agreement with quantum scattering results. Notably, it performed well for a heavy-light-heavy reaction, even though it does not explicitly account for corner-cutting effects. Recent extensions to SCTST that improve its treatment of deep tunneling were also evaluated within the reduced-dimensionality framework. The success of RD SCTST in this study suggests its potential applicability to larger systems.


Page 3 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Quantum mechanical methods for calculating reaction rate constants normally involve constructing a potential energy surface (PES) from ab initio energy calculations and then performing either time-dependent or time-independent scattering calculations on this PES.1-10 For reactions involving more than a few atoms, constructing a PES for all internal degrees of freedom (DOFs) of the system and performing scattering calculations are often prohibitively expensive in terms of computational effort.10 To our knowledge, the largest systems to which these methods have been applied, with full treatment of all DOFs, are the reactions H + CH4 à H2 + CH311 and H2 + CH3 à H + CH4 (and isotopologues).12 Semiclassical Transition State Theory (SCTST), developed originally by Miller and colleagues,13-17 is a computationally more efficient method for calculating reaction rate constants that is more tractable for larger systems. In its original form, it is based on second-order vibrational perturbation theory (VPT2),18-19 which requires only a fourth-order expansion of the PES about the transition state, although SCTST can also accommodate higher-order perturbative approaches.17 As VPT2 treats anharmonic coupling among all internal degrees of freedom of the system, SCTST relies upon ab initio calculations of vibrational frequencies and anharmonic constants at the transition state. SCTST simulates semiclassical tunneling through a potential barrier and recovers in an approximate manner quantum mechanical effects not included in conventional Transition State Theory (TST). SCTST offers computational advantages over quantum scattering methods because (1) it requires only derivatives of the PES near the transition state, rather than the entire PES, thereby reducing the number of ab initio energy 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 47

calculations required, and (2) rate constant calculations are considerably less computationally intensive in SCTST than in quantum scattering methods. Recently, Wagner20 developed an extension to SCTST that more accurately treats tunneling through the potential barrier at low energies, which enables further improvements in the accuracy of SCTST, particularly at low temperature. SCTST has been applied previously to reactions such as H + H2 à H2 + H,16 HBr + Cl à Br + HCl,21 HO + H2 à H2O + H and isotopologues,22 and Cl + CH4 à HCl + CH3 and isotopologues.23 Calculating the anharmonic constants required by SCTST involves on the order of 3F ab initio energy calculations, where F is the number of internal DOFs of the system. Applying SCTST to larger systems is computationally expensive, and a method with better scaling is desirable. Improvements in the scaling of quantum scattering methods have been achieved with reduced-dimensionality (RD) approaches.8, 24-30 In these approaches, the potential is calculated by ab initio methods as a function of only a subset of the internal DOFs, and the remaining DOFs are treated via their zero-point energy (ZPE) contribution to the potential. Substantially fewer ab initio calculations are therefore required to construct an RD PES than a full dimensionality PES. Quantum scattering calculations have been performed previously on RD PESs for a number of reactions,31-52 and the results suggest that only a subset of DOFs need to be considered explicitly in order to calculate rate constants with reasonable accuracy. In this study, we seek to determine whether treating only a subset of DOFs in SCTST can also yield rate constants with acceptable accuracy. We present a modified version of SCTST, Reduced-Dimensionality Semiclassical Transition State Theory (RD


Page 5 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


SCTST), in which only a subset of DOFs are treated explicitly. We apply RD SCTST to the following H abstraction and exchange reactions of hydrocarbon molecules: H + CH4 à H2 + CH3

(R1)

H + cyc-C3H6 à H2 + cyc-C3H5

(R2)

CH3 + CH4 à CH4 + CH3

(R3)

Rate constants for these reactions have been calculated previously by quantum scattering methods on 2-D PESs.44, 51 We use these 2-D PESs in our RD SCTST calculations to evaluate frequencies and anharmonic constants at the transition state and to calculate barrier heights. Rate constants from RD SCTST are compared to quantum scattering results, thereby enabling a systematic evaluation of the accuracy of RD SCTST with respect to RD quantum scattering methods. R1 and R2, both light-light-heavy reactions, are the simplest H abstraction reactions involving an alkane and a cyclic alkane, respectively, and therefore serve as suitable prototypical reactions for our method. All three reactions involve cleavage of a strong C-H bond53 and are important in hydrocarbon combustion reactions.54 R3 is a heavy-light-heavy reaction with a small skew angle and large reaction path curvature. Corner-cutting effects have been shown to be important for such reactions, meaning that tunneling along paths in the concave region of the PES in between the reactant and product channels contributes significantly to the reaction rate.55-56 We apply the RD SCTST method to R3 in order to determine whether it can replicate the results of quantum scattering calculations, which implicitly include corner-cutting effects. All three reactions involve the transfer of a light H atom, suggesting that quantum mechanical tunneling effects are particularly important in determining rate constants at low



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 47

temperature. Consequently, we expect rate constants calculated using RD SCTST and quantum scattering methods to be considerably greater than those calculated using conventional TST at low temperature. Comparing our RD SCTST results to TST results will enable us to determine the extent to which RD SCTST recovers quantum mechanical effects. In section 2 of this paper, we provide a brief overview of SCTST before discussing the modifications required to apply it within an RD framework. We then discuss briefly the methods used to construct the RD PESs we use in this study and discuss in detail how we calculate the parameters required by SCTST. In section 3, we compare results from our RD SCTST method to quantum scattering and experimental results. Section 4 summarizes the key findings of this study. 2. Theoretical Background 2.1 Semiclassical Transition State Theory 2.1.1 Theoretical Overview In Semiclassical Transition State Theory (SCTST), the cumulative reaction probability (CRP) is calculated from harmonic frequencies and anharmonic constants of the vibrational modes of the transition state. Harmonic frequencies are calculated by diagonalizing the mass-weighted Hessian matrix, evaluated at the transition state. If rovibrational coupling is neglected, anharmonic constants are calculated as follows:15 ℏ!

𝑥!! = !"!! 𝑓!!!! − !

ℏ!

𝑥!" = !!

! !!

𝑓!!"" −

! ! ! ! !!!" !!! !!!! !!! ! ! !! ! !! ! ! ! !

!!!" !!!" ! !!! !! !

+

(1)

! !!!

! ! ! !!!"# !! !!!! !!! ! ! !! !!! ! !!! !! !!! ! !!!

(2) 6

ACS Paragon Plus Environment

Page 7 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


where ωi denotes the harmonic frequency of the ith mode, and fklm and fklmn denote third and fourth vector derivatives, respectively, of the potential with respect to normalized normal mode eigenvectors. As the vibrational frequency of the reaction mode is imaginary, off-diagonal anharmonic constants for the reaction mode are imaginary. The rate constant k(T) is calculated from a Boltzmann-weighted average of the CRP, here denoted N(Ev):23 ‡ ‡ ‡ ! !!"#$% ! !!"# ! !!"!# !

𝑘 𝑇 =!

! ! !!"#$%"

! !! !"# ! ! !!! !! !

!! !

(3)

‡ ‡ where 𝑄! 𝑇 is the partition function for the reactants, and 𝑄!"#$% (𝑇), 𝑄!"# 𝑇 , and ‡ 𝑄!"!# 𝑇 are the translational, rotational, and electronic partition functions for the

transition state. Ev denotes the total energy in all vibrational degrees of freedom (DOFs) of the transition state and is defined with respect to the reactant energy. Ethresh, the reaction’s quantum mechanical threshold energy, is the vibrationally corrected classical energy of reaction if the reaction is endothermic and zero otherwise.56 The CRP is calculated in SCTST as 𝑁 𝐸! =

{!} 𝑃 !

(𝐸! )

(4)

The set {n} consists of the F – 1 vibrational quantum numbers of the non-reactive modes of the transition state, where F is the total number of vibrational DOFs. The sum is calculated over all configurations {n} allowed at energy Ev, so N(Ev) is 0 for Ev less than the total ZPE of all non-reactive modes. The state-dependent reaction probability, P{n}(Ev), for each configuration is calculated as 𝑃 ! 𝐸! = 1 + exp 2𝜃 ! 𝐸!

!!

(5)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 47

θ{n}(Ev) represents a barrier penetration integral. For a general one-dimensional potential barrier V(x), the barrier penetration integral is defined in WKB Theory as57 !

𝜃 𝐸 =ħ

2𝑚 𝑉 𝑥 − 𝐸

! !

d𝑥

(6)

Miller et al.15 give the following analytical expression for the barrier penetration integral along the reaction coordinate for each configuration {n}: 𝜃 ! 𝐸! = 𝜋 ℏ

!! ! ! ! !! !! !!! !!!!! !! !

! !

(7)

!!!!

where the subscript F denotes the reaction mode, and ΔV represents the reaction’s forward barrier height. The quantity E{n} represents the total amount of energy in the nonreactive modes of the system: 𝐸! =

!!! !!! ℏ 𝜔!

!

𝑛! + ! +

!!! !!!

!!! !!! 𝑥!"

!

𝑛! + !

!

𝑛! + !

(8)

Ω{n} represents an effective reaction mode frequency, through which SCTST accounts for the influence of vibrational coupling on the rate constant: 𝛺{!} = Im ℏ 𝜔! +

!!! !!! Im(𝑥!" )

!

𝑛! + !

(9)

The above treatment is analogous to calculating the barrier penetration integral for a symmetric Eckart potential, Vse(s):20 !

𝜃 ! 𝐸! = ℏ

2 𝑉!" 𝑠 − 𝐸! + 𝐸 !

! !

d𝑠

(10)

where s is the mass-weighted reaction coordinate, with units of (mass1/2 length). Vse(s) is defined as follows: 𝑉!" 𝑠 = Δ𝑉 − 𝐷 + 4 𝐷

!"# ! ! !!!"# ! ! !

(11)

with


Page 9 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


!!

𝐷 = − !! !

(12)

!!

and !

𝛼=

!! !!! ! ℏ

(13)

The limits of integration in eq 10 are given by the classical turning points of Vse(s) with respect to the energy (Ev – E{n}). Due to this definition of the reaction coordinate, the factor of m1/2 in eq 6 is not explicitly included in eq 10, as it is implicitly included in the differential ds. 2.1.2 Deep Tunneling Corrections The deep tunneling extension to SCTST proposed by Wagner20 more accurately represents tunneling through the potential barrier at energies close to the quantum mechanical threshold energy. Instead of a symmetric Eckart potential, the barrier is modeled as a 3-part piecewise Eckart potential, constructed such that its second, third, and fourth derivatives at the transition state equal those implied by Ω{n} and xFF, and the potential at s = ±∞ is equal to the reactant and product energies, respectively. As demonstrated by Wagner,20 the corresponding barrier penetration integral, θ{n}(Ev), can be calculated analytically, in analogy to eq 10. P{n}(Ev) is then calculated as follows:56 !!

𝑃 ! 𝐸!

1 + exp 2 𝜃 ! 𝐸! , = 1 − 𝑃 ! 2 Δ𝑉 − 𝐸! , 1,

𝐸! − 𝐸 ! ≤ Δ𝑉 Δ𝑉 ≤ 𝐸! − 𝐸 ! ≤ Δ𝑉 + min Δ𝑉, Δ𝑉! Δ𝑉 + min Δ𝑉, Δ𝑉! < 𝐸! − 𝐸 !

(14)

We applied this method within the RD framework in addition to the SCTST method described in section 2.1.1 in order to gain insight into the importance of deep tunneling for the reactions in this study.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 47

2.2 Reduced-Dimensionality Semiclassical Transition State Theory In this section, we present the theoretical basis for Reduced-Dimensionality SCTST (RD SCTST). DOFs treated explicitly in the calculation of the CRP are referred to as the active coordinates. Vibrational DOFs consisting only of linear combinations of the active coordinates are active modes, and the remaining vibrational DOFs are spectator modes. The active mode with an imaginary frequency is the reaction mode, and the other modes, which have real frequencies, are the transition modes. The CRP in eq 4 is calculated as a sum over the energetically allowed configurations of the transition modes, and the energy Ev denotes only the energy in all active modes. The statedependent reaction probability and barrier penetration integral for each configuration is calculated as described in section 2.1.1, except that only transition modes are considered in the sums in eqs 8 and 9. ‡ The spectator modes are treated via an additional partition function, 𝑄!"#$ 𝑇 , in

the expression for k(T), as in other previous RD studies:40, 44, 51 ‡ ‡ ‡ ‡ ! !!"#$% ! !!"# ! !!"!# ! !!"#$ !

𝑘 𝑇 =!

! ! !!"#$%"

! !! !"# ! ! !!! !! !

!! !

(15)

‡ 𝑄!"#$ 𝑇 is a product of harmonic oscillator partition functions, calculated from

frequencies obtained by projecting motion along the active coordinates out of the Hessian at the transition state and diagonalizing it. The details of this projection method will be discussed further in section 2.3. 2.3 Reduced-Dimensionality Potential Energy Surface Construction Figure 1 presents a schematic representation of a generic hydrogen abstraction reaction from a hydrocarbon molecule, Y + HaZ à YHa + Z. Ha denotes the abstracted H 10 ACS Paragon Plus Environment

Page 11 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


atom, C represents the carbon atom from which it is abstracted, and R1, R2, and R3 represent the remainder of the Z moiety. In the previous studies considered in this paper,44, 51 reduced-dimensionality potential energy surfaces (RD PESs) were constructed as functions of only two active coordinates, the Y-Ha and Ha-C bond distances (r1 and r2, respectively, Figure 1). It is important to note that the RD approach does not necessitate the treatment of these particular internal coordinates and that others have approached RD scattering calculations differently.31-34, 38 Discussion of the choice of active coordinates in RD calculations can be found in ref 58 and references therein. In the construction of the RD PESs considered in this study, ab initio calculations were performed on a grid of fixed r1 and r2. At each grid point, a partial geometry optimization of the remaining internal DOFs was performed, followed by a high-level single point energy calculation and a frequency calculation to determine the Hessian. Motion along the active coordinates was projected out of the Hessian,44-45, 59 and the harmonic vibrational frequencies of the remaining F – 2 spectator modes were calculated from the massweighted Hessian. The total zero-point energy (ZPE) of the spectator modes was calculated from these post-projection frequencies and added to the ab initio energy of the system in order to account for the effect of the spectator modes on the PES. The Jacobi coordinates R and r at each grid point were calculated from the optimized geometry of the system. Typically, R is defined as the distance between the centers of mass of the moieties YHa and Z, and r as the distance between the moieties Y and Ha (Figure 1). For R2, R was defined differently as the distance between the center of mass of the moiety YHa and the C atom from which Ha is abstracted.51 For R3, r is defined as the distance between Ha and the C atom of the reactant CH3 moiety.44 Note 11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 47

that this treatment does not assume that the Y, Ha, and Z moieties are collinear, as all coordinates except for the Y-Ha and Ha-Z bond distances (but including the Y-Ha-Z angle) are optimized. Jacobi coordinates were transformed into hyperspherical coordinates ρ and δ, defined as follows: !! !

𝑅! = 𝜌 cos 𝛿

!

,

!! !

𝑟 ! = 𝜌 sin 𝛿

!

(16)

where 𝑚! =

!! !!! !! !! !!! !!!

! !

, 𝑚! = ! !!!! , 𝑚! = !

!

!! (!! !!! ) !!

, 𝜇 = 𝑚! 𝑚! 𝑚!

! !

(17)

mY and mZ are the masses of the Y and Z moieties, and mH and mC are the masses of H and C atoms. The ZPE-corrected ab initio energies were then fit to a double-Morse potential function, V(ρ, δ), in hyperspherical coordinates. The calculations used to construct the RD PESs considered in this study were performed using the MOLPRO package60 for R2 and the Gaussian 03 package61 for R1 and R3. Geometry optimizations were performed using second-order Møller-Plesset perturbation theory (MP2) with a correlation consistent polarized valence triple-ζ Dunning basis set62 (cc-pVTZ). Single-point energy calculations were performed at the coupled cluster level including single, double, and perturbative triple excitations [CCSD(T)] with the cc-pVTZ basis set, except for R2, in which the CCSD(T)-F12a/ccpVTZ-F12 level of theory was used.51, 63-65 Hessians were calculated at the MP2 level of theory using a cc-pVTZ basis set, and projections were performed using the rectilinear projection method of Lu and Truhlar59 for R2 and the curvilinear projection method for R1 and R3.44-45


Page 13 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.4 Reduced-Dimensionality Energetics Calculations For RD SCTST calculations, the frequencies of the active modes were calculated from the mass-weighted Hessian matrix, evaluated at the transition state of the PES. At the transition states of R1 and R3, the Y, Ha, and Z moieties are collinear,44 and, for R2, they are nearly collinear, with a bond angle of 176.0°.51 For simplicity, we assumed that the transition states of all three reactions are collinear, which allowed us to consider only one dimensional motion of each of the moieties along the Y-Ha-Z axis in constructing the Hessian matrices. Second derivatives of the potential V with respect to the positions xY, xH, and xZ of the moieties Y, Ha, and Z, respectively, are calculated analytically as follows: ! !!! ! !!! !

= −!

!!

!

!

! !!! !"

= −!

!! ! !!!

− !"

!

(18)

!

+ !" !"

(19)

!

!!!

= !!

(20)

The chain rule is applied to calculate derivatives of V(ρ, δ), in hyperspherical coordinates, with respect to the Jacobi coordinates r and R: ! !" ! !"

=

!!

=

!!

!

!

!

sin 𝛿 !" + !

!"# ! !

cos 𝛿 !" −

!

!"

!"# ! ! !

!"

(21)

(22)

The masses mY, mH, and mZ are used to weight the 3 × 3 Hessian matrix before diagonalizing it in order to calculate the frequencies of the reaction mode and transition mode. Anharmonic constants are calculated from derivatives of the potential at the transition state with respect to the normal mode eigenvectors, {vn}: 13 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

! !𝒗𝒏

!

= 𝒗𝒏 ∙

! ! 𝑚! ! !! , !

!

! ! 𝑚!! !! , !

Page 14 of 47

!

! ! 𝑚! ! !! !

(23)

The forward barrier height used in our RD SCTST calculations, ΔVSC‡, is calculated as follows: ‡ Δ𝑉!"

ℏ !! !!! ! ! !

= 𝑉 𝜌!" , 𝛿!" − 𝑉 𝜌!"# , 𝛿! − !

! !

!! !! !! ! !!!!"# , !!!!

(24)

where ρTS and δTS represent the hyperspherical coordinates of the transition state. ρmax denotes the maximum value of ρ considered in the quantum mechanical scattering calculation for each reaction, and δr denotes the value of δ at the bottom of the reactant well of the PES at ρmax, which was calculated numerically. The last term in eq 24 represents an effective zero-point energy associated with vibration of the reactant Ha-Z bond. This definition of ΔVSC‡ includes the ZPE of all reactant vibrational modes, both explicitly via the last term of eq 24 and implicitly via the ZPE of the spectator modes included in the PES. As a result, ZPEs were not included in the reactant vibrational partition functions in eq 15. Additionally, ΔVSC‡ does not include the ZPE of the transition mode at the transition state, as this energy is included in E{n} (eq 8). The reaction’s reverse barrier height, ΔVr‡ was determined similarly: ℏ

Δ𝑉!‡ = 𝑉 𝜌!" , 𝛿!" − 𝑉 𝜌!"# , 𝛿! − !

!! !!!

!! !

! !

!! !! !!! ! !!! !"# , !!!!

(25)

where δp denotes the value of δ at the bottom of the product well of the PES. The derivative with respect to r1, the product bond length, is defined as follows: ! !!!

!

= !" + !

!! ! !!!

! !"

(26)

Ethresh, the reaction’s quantum mechanical threshold energy, can be calculated as follows: ‡ 𝐸!"#$%" = max(0, Δ𝑉!" − Δ𝑉!‡ )

(27) 14


Page 15 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.5 Quantum Scattering Calculations In this study, the accuracy of the RD SCTST method is evaluated with respect to RD quantum scattering calculations, in particular the R-matrix propagation scheme,66 performed on the same two-dimensional PESs. We provide a brief overview of these scattering methods here; more detailed descriptions can be found in previous studies.38-45, 47-50

The Hamiltonian in hyperspherical coordinates is given as follows:67 !

𝐻 = − !!

!! !!!

! !!

!!

!

+ !! !! ! − !!! − !! + 𝑉 𝜌, 𝛿

(28)

The R-matrix propagation scheme is used to solve the time-independent Schrödinger equation and calculate the scattering matrix S. The CRP, N(Ev), is calculated from S as follows: 𝑁(𝐸! ) =

!,!

𝑆!→! 𝐸

!

(29)

where the subscripts i and f denote the initial and final states of the reactants and products, respectively. The quantum scattering calculations performed for R1, R2, and R3 employed the J-shifting approximation to calculate the CRP.25, 30, 68 With this approximation, the thermal rate constant, k(T), can be calculated using eq 15. 3. Results and Discussion 3.1 H + CH4 à H2 + CH3 (R1) 3.1.1 Energetics Table 1 presents the relevant harmonic frequencies for the vibrational modes of the transition state of R1. The reported pre-projection frequencies were calculated by Banks et al.44 from the full-dimensionality mass-weighted Hessian, and the post-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 47

projection frequencies were calculated after motion along the active coordinates, r1 and r2, was projected out of this Hessian. A comparison of these two sets of frequencies indicates that the active modes correspond to the pre-projection modes with frequencies 1648.6i cm-1 and 1352.9 cm-1 and that projection left the other spectator mode frequencies predominantly unchanged. The frequency of the reaction mode from the RD PES (1782.3i cm-1) is similar to that of the pre-projection reaction mode (1648.6i cm-1). However, the frequency of the transition mode from the RD PES (1352.9 cm-1) differs considerably from that of the corresponding pre-projection mode (1941.4 cm-1). This difference arises from the RD nature of the PES, in particular that it only treats the two active coordinates explicitly. Anharmonic constants for R1 are also presented in Table 1. ________________________________________________________________________ Table 1. Pre- and Post-projection Frequencies, and Frequencies and Anharmonic Constants Calculated from the RD PES for R1 at the Transition Statea Pre-projection frequencies

1648.6i, 543.5 (2), 1076.5, 1131.5 (2), 1457.6 (2), 1941.4, 3127.8, 3289.4 (2)

Post-projection frequencies

543.5 (2), 1047.5, 1131.5 (2), 1457.6 (2), 3125.5, 3289.4 (2)

RD PES frequencies (ωF, ω1)

1782.3i, 1352.9

RD PES anharmonic constants (x11, x1F, xFF) 0.33, −160.59i, −166.44 ________________________________________________________________________ a All frequencies and anharmonic constants are reported in cm-1. Multiplicities are indicated in parentheses. The reported pre- and post-projection frequencies are from Banks et al.44 The two pre-projection frequencies in bold represent those of active modes. The RD PES frequencies and anharmonic constants were calculated from derivatives of the RD PES at the transition state, as described in section 2.4. The subscript 1 denotes the transition mode, and the subscript F denotes the reaction mode. Forward barrier heights for R1 are presented in Table 2. ΔV‡ denotes the barrier height without any ZPE contributions, and ΔVa‡ denotes the vibrationally adiabatic barrier height. Both of these quantities were calculated directly using ab initio methods. 16 ACS Paragon Plus Environment

Page 17 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ΔVSC,a‡ represents the adiabatic barrier height from the RD SCTST method, i.e. ΔVSC‡ plus the ZPE of the transition mode at the transition state. There is only a small difference between ΔVSC,a‡ and ΔVa‡, which arises mainly because of the frequency differences shown in Table 1. It should be noted that the transition state ZPE includes an anharmonic term but that its contribution to ΔVSC,a‡ is negligible (0.00024 kcal mol-1). ________________________________________________________________________ Table 2. Reaction Energetics for R1a ΔV‡ 15.149 ‡ ΔVa 13.785 ΔVSC,a‡ 13.509

____________________________________________________________________________________________________________ a

ΔV‡ and ΔVa‡ represent the forward barrier height and vibrationally adiabatic forward barrier height, respectively, as calculated by Banks et al.44 using ab initio methods. ΔVSC,a‡ represents the barrier height used in RD SCTST calculations, as defined in eq 24, plus the harmonic and anharmonic ZPEs of the transition state. All barrier heights are reported in kcal mol-1. Figure 2a presents the potential barrier along the minimum-energy path (MEP) of the PES for R1, calculated by methods described in Truhlar et al.,56 and the semiclassical approximation to the barrier (eq 11). In order to allow comparison between these barriers, the semiclassical barrier was constructed for Ω{n} equal to |ħ ωF|. Figure 2b presents the WKB barrier penetration integrals for each of these barriers, calculated as a function of energy using eq 6. At low energies, the MEP barrier is wider than the semiclassical barrier, resulting in a greater integral for the MEP barrier. The widths of the barriers are more similar near the transition state, so the integrals converge as E increases. It is important to note that integrals presented in Figure 2b cannot be used directly to calculate reaction probabilities because they do not include coupling effects. Nevertheless, comparing them provides an idea of the relative amount of quantum mechanical tunneling associated with each barrier. 17 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 47

3.1.2 Rate Constants Calculated rate constants for R1 are presented in Figure 3a. The RD SCTST method yields rate constants in good agreement with the quantum scattering results from Banks et al.44 between 250 K and 1000 K, differing at most by a factor of 2.64 at 250 K. These RD SCTST rate constants were calculated without the deep tunneling corrections proposed by Wagner20 (section 2.1.2). The effect of these corrections will be discussed in section 3.4. The ratio of the RD SCTST rate constant to the scattering rate constant is presented in the inset in Figure 3a. This ratio decreases as T increases from 250 K to 800 K but increases from 800 K to 1000 K. However, it remains between 1 and 2.64 over this T range. It is worth reiterating that the same PES was used to calculate the rate constants in both methods, so discrepancies between the RD SCTST and quantum scattering rate constants arise from differences in dynamics calculations rather than those in the underlying potential energy calculations. The RD SCTST and scattering rate constants at 250 K are 488 and 185 times greater than the rate constant calculated using conventional TST, respectively, reflecting the inclusion of quantum mechanical tunneling effects in the RD SCTST method. Additionally, RD SCTST results exhibit good agreement with experimentally determined rate constants69 (Figure 3b). Rate constants calculated by all methods discussed in this study and experimentally determined rate constants are presented in Table 3.


Page 19 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


________________________________________________________________________ Table 3. Rate Constants for R1 in cm3 molecule-1 s-1, Calculated by the Methods Described in this Study and Determined Experimentallya T (K) TST RD SCTST scatteringb RD SCTST-DT exptlc 200 8.52(−26) 2.86(−21) 5.67(−22) 2.12(−21) 250 8.26(−23) 4.03(−20) 1.53(−20) 3.56(−20) 300 8.25(−21) 4.40(−19) 2.33(−19) 4.20(−19) 350 2.26(−19) 3.60(−18) 2.25(−18) 3.53(−18) 2.47(−18) 400 2.77(−18) 2.17(−17) 1.50(−17) 2.15(−17) 1.82(−17) 450 1.98(−17) 9.91(−17) 7.34(−17) 9.87(−17) 8.96(−17) 500 9.73(−17) 3.60(−16) 2.80(−16) 3.60(−16) 3.33(−16) 600 1.11(−15) 2.82(−15) 2.34(−15) 2.83(−15) 2.57(−15) 700 6.57(−15) 1.36(−14) 1.16(−14) 1.37(−14) 1.19(−14) 800 2.59(−14) 4.71(−14) 4.07(−14) 4.74(−14) 3.99(−14) 1000 1.90(−13) 3.01(−13) 2.57(−13) 3.02(−13) 1300 1.34(−12) 1.94(−12) 1.57(−12) 1.95(−12) 1600 4.92(−12) 6.91(−12) 5.24(−12) 6.95(−12) 2000 1.65(−11) 2.28(−11) 1.57(−11) 2.29(−11) ________________________________________________________________________ a Numbers in parentheses denote powers of 10. “RD SCTST-DT” denotes the RD SCTST method applied with the deep tunneling corrections discussed in section 2.1.2. b From ref. 44. c From ref. 69. The RD SCTST rate constant is marginally greater than the scattering rate constant over the range of temperatures shown in Figure 3a, although they converge as T increases to 1000 K. This behavior arises from differences in the cumulative reaction probability (CRP) calculated in both methods. The CRP is multiplied by a Boltzmann factor and integrated with respect to energy in order to calculate the rate constant (eq 15). Consequently, the CRP at lower energies contributes more to this integral than the CRP at greater energies. At 250 K, the integral is determined primarily by the CRP at energies less than approximately 0.6 eV. Because the CRP from the RD SCTST method is greater than that from the quantum scattering method at these energies (Figure 4a), the Boltzmann-weighted CRP is also greater, and the RD SCTST method yields a greater rate constant at 250 K. At energies near ΔVSC,a‡ (indicated by vertical dashed lines in Figure 19 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 47

4), the CRPs of the two methods are closer in value. The Boltzmann factor emphasizes this energy range more as T increases from 250 K to 800 K. Consequently, the Boltzmann-weighted CRPs from the two methods, and therefore the rate constants, become closer, as indicated by a decrease in the ratio of the rate constants from 250 K to 800 K (Figure 4b, Figure 3a inset). At higher energies, the quantum scattering CRP is less than the RD SCTST CRP. This range of the CRP contributes more to the integral as T increases further, which explains the increase in the ratio of the RD SCTST and scattering rate constants from 800 K to 1000 K. 3.2 H + cyc-C3H6 à H2 + cyc-C3H5 (R2) The frequencies and anharmonic constants used in RD SCTST calculations for R2 are presented in Table 4. As in R1, the reaction mode pre-projection frequency (1342.30i cm-1) differs somewhat from that obtained from the RD PES (1527.64i cm-1), and there is a greater difference between the transition mode frequencies (2255.22 cm-1 vs. 1537.51 cm-1). This contributes to a difference between the vibrationally adiabatic barrier height obtained from ab initio calculations (ΔVa‡, 13.47 kcal mol-1) and vibrationally adiabatic semiclassical barrier height (ΔVSC,a‡, 12.07 kcal mol-1). The anharmonic ZPE of the transition mode contributes more to ΔVSC,a‡ (0.08 kcal mol-1) than in R1.


Page 21 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


________________________________________________________________________ Table 4. Energetic Parameters, Frequencies, and Anharmonic Constants for R2a Pre-projection frequencies 1342.30i, 2255.22 RD PES frequencies 1527.64i, 1537.51 RD PES anharmonic constants (x11, x1F, xFF) 109.75, −618.86i, −254.58 ΔV‡ 14.56 ΔVa‡ 13.47 ‡ ΔVSC,a 12.07 ________________________________________________________________________ a The reported pre-projection frequencies correspond to those of the active modes. ΔV‡, ΔVa‡, and ΔVSC,a‡ are defined as in Table 2. Frequencies and anharmonic constants are reported in cm-1, and barrier heights are reported in kcal mol-1. Potential barriers for R2 are presented in Figure 5a. Although the semiclassical barrier has a width similar to the MEP barrier near the top, their shapes differ significantly. The reactant and product asymptotes of the semiclassical barrier are significantly greater than the quantum mechanical threshold energy. Nevertheless, the rate constants calculated by RD SCTST (without deep tunneling corrections) exhibit very close agreement with quantum scattering rate constants (Figure 5b). At 250 K, the RD SCTST rate constant is 1.73 times greater than the scattering rate constant and 95 times greater than the TST rate constant. The RD SCTST and scattering rate constants converge at high T. Additionally, the RD SCTST rate constant agrees well with experimental results53, 70 from 250 K to 1000 K (Figure 5c). Calculated and experimentally determined rate constants for R2 are presented in Table 5.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 47

________________________________________________________________________ Table 5. Rate Constants for R2 in cm3 molecule-1 s-1, Calculated by the Methods Described in this Study and Determined Experimentallya T (K) TST RD SCTST scatteringb RD SCTST-DT exptl 360 6.49(−19) 1.04(−17) 8.08(−18) 1.08(−17) 2.05(−17)c 390 2.90(−18) 3.52(−17) 2.83(−17) 3.65(−17) 7.25(−17)c 420 1.05(−17) 1.02(−16) 8.38(−17) 1.05(−16) 2.14(−16)c 450 3.24(−17) 2.60(−16) 2.17(−16) 2.67(−16) 5.45(−16)c 480 8.73(−17) 5.96(−16) 5.02(−16) 6.10(−16) 1.24(−15)c 510 2.10(−16) 1.25(−15) 1.06(−15) 1.28(−15) 2.55(−15)c 540 4.63(−16) 2.44(−15) 2.08(−15) 2.49(−15) 4.84(−15)c 630 3.21(−15) 1.28(−14) 1.09(−14) 1.31(−14) 6.23(−15)d 660 5.49(−15) 2.04(−14) 1.73(−14) 2.08(−14) 9.50(−15)d 690 8.99(−15) 3.13(−14) 2.64(−14) 3.19(−14) 1.40(−14)d 720 1.42(−14) 4.65(−14) 3.91(−14) 4.73(−14) 1.99(−14)d 750 2.16(−14) 6.73(−14) 5.63(−14) 6.84(−14) 2.75(−14)d ________________________________________________________________________ a Numbers in parentheses denote powers of 10. Methods for calculating rate constants are denoted as in Table 3. b From ref. 51. c From ref. 53. d From ref. 70. 3.3 CH3 + CH4 à CH4 + CH3 (R3) Frequencies and anharmonic constants for R3 are presented in Table 6. The frequencies of both active modes as calculated from the RD PES differ somewhat from the corresponding pre-projection frequencies (1683.55i cm-1 vs. 1948.8i cm-1 for the reaction mode, and 494.74 cm-1 vs. 540.8 cm-1 for the transition mode). Additionally, the vibrationally adiabatic barrier height from ab initio calculations (ΔVa‡) is 0.691 kcal mol-1 greater than that calculated from the RD PES (ΔVSC,a‡).


Page 23 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


________________________________________________________________________ Table 6. Energetic Parameters, Frequencies, and Anharmonic Constants for R3a Pre-projection frequencies 1949.8i, 540.8 RD PES frequencies 1683.55i, 494.74 RD PES anharmonic constants (x11, x1F, xFF) −1.64, −32.78i, −109.95 ΔV‡ 18.142 ΔVa‡ 17.705 ‡ ΔVSC,a 16.999 ________________________________________________________________________ a The parameters reported here are defined as in Table 2. Frequencies and anharmonic constants are reported in cm-1, and barrier heights are reported in kcal mol-1. Figure 6a shows the potential barrier along the MEP of the RD PES and the symmetric Eckart potential representing the potential barrier in the RD SCTST method. The two barriers have very similar shapes near the transition state, but the MEP barrier is significantly wider than the semiclassical barrier at low energies. Nevertheless, the RD SCTST method (without deep tunneling corrections) yielded rate constants in good agreement with scattering results from Banks et al.44 (Figure 6b). The RD SCTST rate constant at 250 K is 2.29 times greater than the scattering rate constant and 4358 times greater than the TST rate constant. The RD SCTST and scattering rate constants converge at high T, although the Arrhenius plot of the RD SCTST rate constant exhibits more curvature than that of the scattering rate constant. Figure 6c presents our RD SCTST rate constants in comparison with experimental results.71-72 Rate constants for R3 are presented in Table 7.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 47

________________________________________________________________________ Table 7. Rate Constants for R3 in cm3 molecule-1 s-1, Calculated by the Methods Described in this Study and Determined Experimentallya T (K) TST RD SCTST scatteringb RD SCTST-DT exptl 200 1.68(−31) 3.63(−25) 1.83(−26) 1.75(−25) 1.46(−28)c 250 1.11(−27) 4.85(−24) 2.12(−24) 3.71(−24) 2.17(−25)c 300 4.08(−25) 7.64(−23) 8.90(−23) 7.01(−23) 2.82(−23)c 350 2.88(−23) 1.07(−21) 1.74(−21) 1.04(−21) 400 7.24(−22) 1.10(−20) 1.91(−20) 1.09(−20) 1.45(−20)d 450 9.20(−21) 8.06(−20) 1.36(−19) 8.03(−20) 1.03(−19)d 500 7.23(−20) 4.39(−19) 7.02(−19) 4.39(−19) 4.94(−19)d 600 1.71(−18) 6.61(−18) 9.30(−18) 6.62(−18) 5.19(−18)d 700 1.76(−17) 5.24(−17) 6.51(−17) 5.25(−17) 2.78(−17)d 800 1.07(−16) 2.70(−16) 2.99(−16) 2.71(−16) 9.80(−17)d 1000 1.52(−15) 3.15(−15) 2.83(−15) 3.16(−15) 1300 2.15(−14) 3.84(−14) 2.65(−14) 3.86(−14) 1600 1.31(−13) 2.17(−13) 1.21(−13) 2.18(−13) 2000 7.23(−13) 1.14(−12) 5.11(−13) 1.14(−12) ________________________________________________________________________ a Numbers in parentheses denote powers of 10. Methods for calculating rate constants are denoted as in Table 3. b From ref. 44. c From ref. 72. d From ref. 71. For reactions with small skew angles and large reaction path curvature such as R3, corner-cutting effects have been shown to be important.55-56, 73-77 Corner-cutting involves tunneling along paths between the reactant and product channels of the PES other than the MEP. The barrier along a chosen corner-cutting path is narrower than that along the MEP, such that the associated barrier penetration integral is less than that of the MEP, even though the barrier is higher. One possible corner-cutting path on the RD PES for R3 is presented in Figure 7a as the blue line (“CC path”). The potential barrier associated with this path is presented in Figure 7b (blue curve) in comparison to the MEP barrier (red curve). Although the corner-cutting barrier is higher than the MEP barrier, it is considerably narrower, and it is also similar in width to the semiclassical barrier at lower energies (black dashed curve). The RD SCTST method does not include corner24 ACS Paragon Plus Environment

Page 25 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cutting effects, as it does not consider the concave region of the PES in between the reactant and product channels, but it nonetheless yields acceptable rate constants because it assumes a sufficiently narrow potential barrier. Further investigation is needed to determine whether the narrowness of the semiclassical barrier relative to the MEP barrier properly compensates for the lack of corner-cutting effects in all reactions with small skew angles and large reaction path curvatures. These findings warrant further investigation and the application of RD SCTST to other reactions with different reaction path curvatures. 3.4 Rate Constants with Deep Tunneling Corrections In order to evaluate the importance of the deep tunneling corrections to the SCTST method proposed by Wagner20 (section 2.1.2) within the RD framework, we compared RD SCTST rate constants calculated with and without these corrections. We refer to RD SCTST applied with these corrections as “RD SCTST-DT.” The ratios of the RD SCTST-DT rate constants to the RD SCTST rate constants are presented in Figure 8a for R1, R2, and R3. Numerical values of these rate constants are presented in Tables 3, 5, and 7. For R1 and R2, the RD SCTST-DT rate constant differed from the RD SCTST rate constant by no more than 16% from 250 K to 1000 K. Differences were greater for R3, but the RD SCTST-DT rate constant was nonetheless only 24% less than the RD SCTST rate constant at 250 K. The RD SCTST-DT potential barrier was wider than the RD SCTST barrier at low energies for R3 (Figure 8b). This difference contributed more to a difference in rate constants at low T for R3 because its potential barrier height is significantly less than those of R1 and R2. These results indicate that, for the temperature range and reactions considered in this study, applying deep tunneling corrections does not 25 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 47

significantly improve the accuracy of RD SCTST rate constant calculations with respect to quantum scattering calculations. 4. Conclusions We present a Reduced-Dimensionality Semiclassical Transition State Theory (RD SCTST) for calculating reaction rate constants. It offers computational advantages over RD quantum mechanical scattering methods and full-dimensionality SCTST. RD potential energy surfaces (PESs) developed previously for three H abstraction and exchange reactions of hydrocarbon molecules were used to calculate the harmonic frequencies and barrier heights required by RD SCTST. These parameters exhibited acceptable agreement with those obtained from ab initio calculations. Rate constants for the three reactions calculated using RD SCTST were compared to those from quantum scattering calculations performed on the same PESs, thereby enabling a systematic evaluation of RD SCTST. We observed good agreement between these rate constants from 250 K to 1000 K. RD SCTST results also agreed with experimental results for all three reactions. RD SCTST performed well also for R3, even though it does not account explicitly for the corner-cutting effects shown previously to be important for heavy-lightheavy reactions. Additionally, we applied deep tunneling extensions to SCTST within the RD framework and found that they did not appreciably improve the accuracy of calculated rate constants over the temperature range considered in this study. The results suggest that the RD SCTST method has significant potential for calculating reaction rate constants with minimal computational expense.


Page 27 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Acknowledgements X. Shan and D. Clary thank the financial support of the Leverhulme Trust Project (Grant No. RPG-2013-321). S. Greene gratefully acknowledges financial support from the Rhodes Trust through a Rhodes Scholarship for graduate study.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 47

References 1.

Tal-‐Ezer, H.; Kosloff, R., An Accurate and Efficient Scheme for Propagating the Time

Dependent Schrödinger Equation. J. Chem. Phys. 1984, 81, 3967–3971. 2.

Mowrey, R. C.; Kouri, D. J., Close-‐Coupling Wave Packet Approach to Numerically

Exact Molecule–Surface Scattering Calculations. J. Chem. Phys. 1986, 84, 6466–6473. 3.

Neuhauser, D.; Baer, M., The Application of Wave Packets to Reactive Atom–Diatom

Systems: A New Approach. J. Chem. Phys. 1989, 91, 4651–4657. 4.

Miller, W. H., Recent Advances in Quantum Mechanical Reactive Scattering Theory,

Including Comparison of Recent Experiments with Rigorous Calculations of State-to-State Cross Sections for the H/D+H2→H2/HD+H Reactions. Annu. Rev. Phys. Chem. 1990, 41, 245–281. 5.

Neuhauser, D.; Baer, M.; Judson, R. S.; Kouri, D. J., The Application of Time-Dependent

Wavepacket Methods to Reactive Scattering. Comput. Phys. Commun. 1991, 63, 460–481. 6.

Jackson, B., Time-Dependent Wave Packet Approach to Quantum Reactive Scattering.

Annu. Rev. Phys. Chem. 1995, 46, 251–274. 7.

Balakrishnan, N.; Kalyanaraman, C.; Sathyamurthy, N., Time-Dependent Quantum

Mechanical Approach to Reactive Scattering and Related Processes. Physics Reports 1997, 280, 79–144. 8.

Nyman, G.; Yu, H.-G., Quantum Theory of Bimolecular Chemical Reactions. Rep. Prog.

Phys. 2000, 63, 1001–1059. 9.

Clary, D. C.; Kroes, G.-J., Chemical Reactions. In Scattering: Scattering and Inverse

Scattering in Pure and Applied Science, Pike, E. R.; Sabatier, P. C., Eds. Academic Press: London, 2001; pp 1068–1080.

28


Page 29 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


10.

Althorpe, S. C.; Clary, D. C., Quantum Scattering Calculations on Chemical Reactions.

Annu. Rev. Phys. Chem. 2003, 54, 493–529. 11.

Wang, D., A Time-Dependent Quantum Dynamics Study of the H2 +CH3 → H + CH4

Reaction. J. Chem. Phys. 2002, 117, 9806–9810. 12.

Nyman, G.; van Harrevelt, R.; Manthe, U., Thermochemistry and Accurate Quantum

Reaction Rate Calculations for H2/HD/D2 + CH3. J. Phys. Chem. A 2007, 111, 10331–10337. 13.

Miller, W. H., Semiclassical Limit of Quantum Mechanical Transition State Theory for

Nonseparable Systems. J. Chem. Phys. 1975, 62, 1899–1906. 14.

Miller, W. H., Semi-Classical Theory for Non-Separable Systems: Construction of

"Good" Action-Angle Variables for Reaction Rate Constants. Faraday Discuss. Chem. Soc. 1977, 62, 40–46. 15.

Miller, W. H.; Hernandez, R.; Handy, N. C.; Jayatilaka, D.; Willetts, A., Ab Initio

Calculation of Anharmonic Constants for a Transition State, with Application to Semiclassical Transition State Tunneling Probabilities. Chem. Phys. Lett. 1990, 172, 62–68. 16.

Cohen, M. J.; Handy, N. C.; Hernandez, R.; Miller, W. H., Cumulative Reaction

Probabilities for H + H2 → H2 + H from a Knowledge of the Anharmonic Force Field. Chem. Phys. Lett. 1992, 192, 407–416. 17.

Hernandez, R.; Miller, W. H., Semiclassical Transition State Theory: A New Perspective.

Chem. Phys. Lett. 1993, 214, 129–136. 18.

Mills, I. M., Vibration-Rotation Structure in Asymmetric- and Symmetric-Top

Molecules. In Molecular Spectroscopy, Mathews, K. N. R. W., Ed. Academic Press: 1972; pp 115–140.

29



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

19.

Page 30 of 47

Barone, V., Anharmonic Vibrational Properties by a Fully Automated Second-Order

Perturbative Approach. J. Chem. Phys. 2005, 122, 14108. 20.

Wagner, A. F., Improved Multidimensional Semiclassical Tunneling Theory. J. Phys.

Chem. A 2013, 117, 13089–13100. 21.

Cohen, M. J.; Willetts, A.; Handy, N. C., Reaction Rates of BrH + Cl → Br + HCl Using

Semiclassical Transition State Theory. Chem. Phys. Lett. 1994, 223, 459–464. 22.

Nguyen, T. L.; Stanton, J. F.; Barker, J. R., Ab Initio Reaction Rate Constants Computed

Using Semiclassical Transition-State Theory: HO + H2 → H2O + H and Isotopologues. J. Phys. Chem. A 2011, 115, 5118–5126. 23.

Barker, J. R.; Nguyen, T. L.; Stanton, J. F., Kinetic Isotope Effects for Cl + CH4 ⇋ HCl +

CH3 Calculated Using Ab Initio Semiclassical Transition State Theory. J. Phys. Chem. A 2012, 116, 6408–6419. 24.

Bowman, J. M., Reduced Dimensionality Theories of Quantum Reactive Scattering. In

Advances in Chemical Physics, Prigogine, I.; Rice, S. A., Eds. John Wiley & Sons, Inc.: Hoboken, NJ, 1985; Vol. 61, pp 115–167. 25.

Bowman, J. M., Reduced Dimensionality Theory of Quantum Reactive Scattering. J.

Phys. Chem. 1991, 95, 4960–4968. 26.

Clary, D. C., Quantum Theory of Chemical Reaction Dynamics. Science 1998, 279,

1879–1882. 27.

Aguillon, F.; Sizun, M., Reduced Dimensionality Wave Packet Study of the NH3+ + H2,

D2 Reaction. J. Chem. Phys. 2000, 112, 10179–10191. 28.

Szichman, H.; Gilibert, M.; González, M.; Giménez, X.; Navarro, A. A., Four-

Dimensional Quantum Mechanical Treatment of Penta-Atomic Systems: Case H2 + C2H → H + 30


Page 31 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


C2H2; Total Reactive Probabilities, Cross Sections, and Rate Constants. J. Chem. Phys. 2000, 113, 176–183. 29.

Szichman, H.; Gilibert, M.; González, M.; Giménez, X.; Aguilar, A., A Four-

Dimensional Quantum Mechanical State-to-State Study of the H2 + C2H → H + C2H2 Reaction. J. Chem. Phys. 2001, 114, 9882–9894. 30.

Bowman, J. M., Overview of Reduced Dimensionality Quantum Approaches to Reactive

Scattering. Theor. Chem. Acc. 2002, 108, 125–133. 31.

Nyman, G.; Clary, D. C., Quantum Scattering Calculations on the CH4 + OH → CH3 +

H2O Reaction. J. Chem. Phys. 1994, 101, 5756–5771. 32.

Nyman, G.; Clary, D. C.; Levine, R. D., Potential Energy Surface Effects on Differential

Cross Sections for Polyatomic Reactions. Chem. Phys. 1995, 191, 223–233. 33.

Nyman, G., Quantum Scattering Calculations on the NH3 + OH → NH2 + H2O Reaction.

J. Chem. Phys. 1996, 104, 6154–6167. 34.

Takayanagi, T., Reduced Dimensionality Calculations of Quantum Reactive Scattering

for the H + CH4 → H2 + CH3 Reaction. J. Chem. Phys. 1996, 104, 2237–2242. 35.

Clary, D. C.; Palma, J., Quantum Dynamics of the Walden Inversion Reaction Cl− +

CH3Cl → ClCH3+Cl−. J. Chem. Phys. 1997, 106, 575–583. 36.

Le Garrec, J.-L.; Rowe, B. R.; Queffelec, J. L.; Mitchell, J. B. A.; Clary, D. C.,

Temperature Dependence of the Rate Constant for the Cl− + CH3Br Reaction Down to 23 K. J. Chem. Phys. 1997, 107, 1021–1024. 37.

Kerkeni, B.; Clary, D. C., Quantum Dynamics and Kinetics of the Abstraction Reactions

by H Atoms of Primary and Secondary Hydrogens in C3H8. Mol. Phys. 2005, 103, 1745–1755.

31



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

38.

Page 32 of 47

Kerkeni, B.; Clary, D. C., The Effect of the Torsional and Stretching Vibrations of C2H6

on the H + C2H6 → H2 + C2H5 Reaction. J. Chem. Phys. 2005, 123, 64305. 39.

Kerkeni, B.; Clary, D. C., Ab Initio Rate Constants from Hyperspherical Quantum

Scattering: Application to H + C2H6 and H + CH3OH. J. Chem. Phys. 2004, 121, 6809–6821. 40.

Kerkeni, B.; Clary, D. C., Ab Initio Rate Constants from Hyperspherical Quantum

Scattering: Application to H + CH4 → H2 + CH3. J. Chem. Phys. 2004, 120, 2308–2318. 41.

Kerkeni, B.; Clary, D. C., Quantum Reactive Scattering of H + Hydrocarbon Reactions.

Phys. Chem. Chem. Phys. 2006, 8, 917–925. 42.

Kerkeni, B.; Clary, D. C., Quantum Scattering Study of the Abstraction Reactions of H

Atoms from CH3NH2. Chem. Phys. Lett. 2007, 438, 1–7. 43.

Banks, S. T.; Clary, D. C., Reduced Dimensionality Quantum Dynamics of Cl + CH4 →

HCl + CH3 on an Ab Initio Potential. Phys. Chem. Chem. Phys. 2007, 9, 933–943. 44.

Banks, S. T.; Tautermann, C. S.; Remmert, S. M.; Clary, D. C., An Improved Treatment

of Spectator Mode Vibrations in Reduced Dimensional Quantum Dynamics: Application to the Hydrogen Abstraction Reactions µ + CH4, H + CH4, D + CH4, and CH3 + CH4. J. Chem. Phys. 2009, 131, 044111. 45.

Banks, S. T.; Clary, D. C., Chemical Reaction Surface Vibrational Frequencies Evaluated

in Curvilinear Internal Coordinates: Application to H + CH4 ⇋ H2 + CH3. J. Chem. Phys. 2009, 130, 024106. 46.

Remmert, S. M.; Banks, S. T.; Clary, D. C., Reduced Dimensionality Quantum Dynamics

of CH3 + CH4 → CH4 + CH3: Symmetric Hydrogen Exchange on an Ab Initio Potential. J. Phys. Chem. A 2009, 113, 4255–4264.

32


Page 33 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


47.

Remmert, S. M.; Banks, S. T.; Harvey, J. N.; Orr-Ewing, A. J.; Clary, D. C., Reduced

Dimensionality Spin-Orbit Dynamics of CH3 + HCl ⇋ CH4 + Cl on Ab Initio Surfaces. J. Chem. Phys. 2011, 134, 204311. 48.

Shan, X.; Clary, D. C., A Reduced Dimensionality Quantum Mechanical Study of the H

+ HCF3 ↔ H2 + CF3 Reaction. Phys. Chem. Chem. Phys. 2013, 15, 18530–18538. 49.

Shan, X.; Remmert, S. M.; Clary, D. C.; Zhang, B.; Liu, K., Crossed-Beam and Reduced

Dimensionality Studies of the State-to-State Integral Cross Sections of the Cl + HCD3(ν) → HCl(ν′) + CD3 Reaction. Chem. Phys. Lett. 2013, 587, 88–92. 50.

Shan, X.; Clary, D. C., Quantum Effects in the Abstraction Reaction by H Atoms of

Primary and Secondary Hydrogens in n-C4H10: A Test of a New Potential Energy Surface Construction Method. Phys. Chem. Chem. Phys. 2013, 15, 1222–1231. 51.

Shan, X.; Clary, D. C., Quantum Dynamics of the Abstraction Reaction of H with

Cyclopropane. J. Phys. Chem. A 2014, 118, 10134–10143. 52.

Takayanagi, T., Reduced-Dimensionality Quantum Reactive Scattering Calculations of

the C(3P) + C2H2 Reaction on a New Potential Energy Surface. Chem. Phys. 2005, 312, 61–67. 53.

Marshall, R. M.; Purnell, H.; Sheppard, A., Reaction of Hydrogen Atoms with

Cyclopropane in the Temperature Range 358-550 K. J. Chem. Soc., Faraday Trans. 2 1986, 82, 929–935. 54.

Arnold, P. A.; Carpenter, B. K., Computational Studies on the Ring Openings of

Cyclopropyl Radical and Cyclopropyl Cation. Chem. Phys. Lett. 2000, 328, 90–96. 55.

Bowman, J. L.; Wagner, A. F., Reduced Dimensionality Theories of Quantum Reactive

Scattering: Applications to Mu + H2, H + H2, O(3P) + H2, D2 and HD. In The Theory of Chemical

33



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 47

Reaction Dynamics, Clary, D. C., Ed. D. Reidel: Dordrecht, Netherlands, 1986; Vol. 170, pp 47– 76. 56.

Truhlar, D. G.; Isaacson, A. D.; Garrett, B. C., Generalized Transition State Theory. In

Theory of Chemical Reaction Dynamics, Baer, M., Ed. CRC Press: Boca Raton, FL, 1985; Vol. 4, pp 65–137. 57.

Griffiths, D. J., Introduction to Quantum Mechanics, 2nd ed.; Pearson Education: Upper

Saddle River, NJ, 2005. 58.

Bowman, J. M.; Schatz, G. C., Theoretical Studies of Polyatomic Bimolecular Reaction

Dynamics. Annu. Rev. Phys. Chem. 1995, 46, 169–196. 59.

Lu, D.-H.; Truhlar, D. G., Bond-‐Distance and Bond-‐Angle Constraints in Reaction-‐Path

Dynamics Calculations. J. Chem. Phys. 1993, 99, 2723–2738. 60.

Werner, H. J., et al. MOLPRO, a Package of Ab Initio Programs, version 2010.2.

61.

Frisch, M. J., et al. Gaussian 03, Revision C.02; Wallingford, CT, 2004.

62.

Dunning, T. H., Gaussian Basis Sets for Use in Correlated Molecular Calculations. I. The

Atoms Boron through Neon and Hydrogen. J. Chem. Phys. 1989, 90, 1007–1023. 63.

Adler, T. B.; Knizia, G.; Werner, H.-J., A Simple and Efficient Ccsd(T)-F12

Approximation. J. Chem. Phys. 2007, 127, 221106. 64.

Knizia, G.; Adler, T. B.; Werner, H.-J., Simplified Ccsd(T)-F12 Methods: Theory and

Benchmarks. J. Chem. Phys. 2009, 130, 054104. 65.

Peterson, K. A.; Adler, T. B.; Werner, H.-J., Systematically Convergent Basis Sets for

Explicitly Correlated Wavefunctions: The Atoms H, He, B–Ne, and Al–Ar. J. Chem. Phys. 2008, 128, 084102.

34


Page 35 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


66.

Stechel, E. B.; Walker, R. B.; Light, J. C., R-Matrix Solution of Coupled Equations for

Inelastic Scattering. J. Chem. Phys. 1978, 69, 3518–3531. 67.

Nyman, G., 2D and 3D Quantum Scattering Calculations on the CH4 + OH → CH3 +

H2O Reaction. Chem. Phys. Lett. 1995, 240, 571–577. 68.

Bowman, J. M., Approximate Time Independent Methods for Polyatomic Reactions. In

Reaction and Molecular Dynamics, Laganà, A.; Riganelli, A., Eds. Springer: New York, 2000; Vol. 75, pp 101–114. 69.

Sutherland, J. W.; Su, M. C.; Michael, J. V., Rate Constants for H + CH4, CH3 + H2, and

CH4 Dissociation at High Temperature. Int. J. Chem. Kinet. 2001, 33, 669–684. 70.

Marshall, R. M.; Purnell, H.; Satchell, P. W., Reaction of Hydrogen Atoms with

Cyclopropane in the Temperature Range from 628 to 779 K. J. Chem. Soc., Faraday Trans. 1 1984, 80, 2395–2403. 71.

Kerr, J. A.; Parsonage, J., Evaluated Kinetic Data on Gas Phase Hydrogen Transfer

Reactions of Methyl Radicals; Butterworths: London, 1976. 72.

Arthur, N. L.; Bell, T. N., An Evaluation of the Kinetic Data for Hydrogen Abstraction

from Silanes in the Gas Phase. Rev. Chem. Intermed. 1978, 2, 37–74. 73.

Garrett, B. C.; Truhlar, D. G.; Grev, R. S.; Magnuson, A. W., Improved Treatment of

Threshold Contributions in Variational Transition-State Theory. J. Phys. Chem. 1980, 84, 1730– 1748. 74.

Truhlar, D. G.; Isaacson, A. D.; Skodje, R. T.; Garrett, B. C., Incorporation of Quantum

Effects in Generalized-Transition-State Theory. J. Phys. Chem. 1982, 86, 2252–2261.

35



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

75.

Page 36 of 47

Isaacson, A. D.; Truhlar, D. G., Polyatomic Canonical Variational Theory for Chemical

Reaction Rates. Separable-Mode Formalism with Application to OH + H2 → H2O + H. J. Chem. Phys. 1982, 76, 1380–1391. 76.

Garrett, B. C.; Truhlar, D. G., Generalized Transition State Theory. Quantum Effects for

Collinear Reactions of Hydrogen Molecules and Isotopically Substituted Hydrogen Molecules. J. Phys. Chem. 1979, 83, 1079–1112. 77.

Meana-Pañeda, R.; Truhlar, D. G.; Fernández-Ramos, A., Least-Action Tunneling

Transmission Coefficient for Polyatomic Reactions. J. Chem. Theory Comput. 2010, 6, 6–17.

36


Page 37 of 47

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figures Z y

x

r1 = r Y

R1

z r2

Ha

C R2 R3

R

Figure 1. A schematic showing the definition of the active internal coordinates r1 and r2 and the Jacobi coordinates R and r for the general H abstraction/exchange reaction Y + HaZ à YHa + Z. HaZ represents a hydrocarbon, with Ha bonded to a C atom in the Z moiety. R1, R2, and R3 represent the remainder of the Z moiety. The white and black circles represent the centers of mass of the YHa and Z moieties, respectively.

37



R1: H + CH 4 H 2 + CH 3

0.020

(a) T.S.

V / a.u.

0.015

0.010

MEP SC approx

0.005

0.000

− 40

− 20 0 20 reaction coordinate / a.u.

40

12 (b)

10 Barrier Penetration Integral

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 47

MEP SC approx 6

3

0 0.000

0.005

0.010 E / a.u.

0.015

0.020

Figure 2. (a) The potential barrier for the H + CH4 à H2 + CH3 reaction (R1) as a function of the reaction coordinate along the minimum-energy path (MEP) of the PES (solid curve), and the symmetric Eckart potential that represents the effective barrier for SCTST calculations (eq 11, dashed curve). In constructing the semiclassical barrier, Ω{n} was set to |ħ ωF| in order to enable comparison with the MEP barrier. (b) Barrier penetration integrals, calculated using eq 6, for the MEP (solid curve) and semiclassical (dashed curve) barriers.

38


Page 39 of 47

R1: H + CH4 750 (a)

500

k RD SCTST / k scatt

2.00

) −1 −1

300

− 16

1.50

1.00 1.0

2.0 3.0 −1 (1000 / T) / K

4.0

3

log ( k / cm molecule s

H2 + CH 3

T/K

− 14

RD SCTST − 18

10

TST

scattering

− 20 1.0

2.0

3.0

4.0

−1

(1000 / T) / K T/K 400

500

300

(b)

−1 −1

)

− 15

− 17

RD SCTST

TST Expt.69

3

log ( k / cm molecule s

− 19

10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


− 21 1.5

2.0

2.5 −1 (1000 / T) / K

3.0

3.5

Figure 3. (a) Rate constants for the H + CH4 à H2 + CH3 reaction (R1) calculated by conventional TST (dashed red line) and by RD SCTST without deep tunneling corrections (solid black curve), presented with scattering results from Banks et al.44 (dashed blue curve). The ratio of the RD SCTST rate constant to the scattering rate constant is plotted in the inset. (b) Experimental results from Sutherland et al.69 presented with TST and RD SCTST results from this study. 39



R1: H + CH 4 H2 + CH 3

N(Ev ) exp (− Ev / kB T ) × 10

11

4.0

(a)

250 K

3.0 RD SCTST 2.0

1.0 0.586 eV scattering 0.0 0.0 1.2

N(Ev) exp (− E v / k BT ) × 10 4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 47

0.2

0.4 E v / eV

0.6

(b)

0.8

800 K scattering RD SCTST

0.8

0.4

0.586 eV 0.0 0.0

0.4

E v / eV

0.8

1.2

Figure 4. The product of the cumulative reaction probability, N(Ev), for the H + CH4 à H2 + CH3 reaction (R1), calculated by RD SCTST and quantum scattering methods, and the Boltzmann factor evaluated at (a) T = 250 K and (b) T = 800 K. The dashed vertical line denotes the vibrationally adiabatic barrier height from the RD SCTST method, ΔVSC,a‡.

40


Page 41 of 47

R2: H + cyc-C3H6

H2 + cyc-C3H5

0.016

(a)

T.S.

V / a.u.

0.012

MEP SC approx

0.008

0.004

0.000 − 40

log10( k / cm 3 molecule− 1s− 1 )

− 12

− 20

750

0 20 40 60 reaction coordinate / a.u. T/K

500

80

300 (b)

− 14

− 16 RD SCTST − 18

TST

− 20

scattering

− 22 1.0

2.0

3.0

(1000 / T) / K

− 12

T/K 500

750

4.0

−1

300 (c)

log10( k / cm 3 molecule− 1s− 1 )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


− 14

− 16

− 18

RD SCTST TST Expt. Marshall (1986) 53 Expt. Marshall (1984) 70

− 20 1.0

1.5

2.0 2.5 (1000 / T) / K − 1

3.0


41


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 42 of 47

Figure 5. (a) The potential barrier for the H + cyc-C3H6 à H2 + cyc-C3H5 reaction (R2) along the MEP of the RD PES (solid curve), and the semiclassical Eckart potential approximation (eq 11, dashed curve). (b) Rate constants for R2 calculated using TST (dashed red line) and RD SCTST (solid black curve), as well as scattering rate constants from Shan and Clary51 (dashed blue curve). (c) RD SCTST and TST rate constants as compared to experimentally determined rate constants.53, 70

42


Page 43 of 47

R3: CH3 + CH4

CH4 + CH3

0.003

(a) MEP SC barrier

T.S.

V / a.u.

0.002

0.001

0.000 − 80

− 40 0 40 reaction coordinate / a.u. 750

500

T/K

80

300 (b)

log10( k / cm 3 molecule− 1s − 1)

− 15

− 18

scattering

− 21

RD SCTST TST

− 24

− 27 1.0

2.0 3.0 −1 (1000 / T) / K 750

500

T/K

4.0

300 (c)

− 15 log10( k / cm 3 molecule− 1s − 1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


− 18

− 21

− 24

RD SCTST TST Expt. Kerr (1976)71 Expt. Arthur (1978)72

− 27 1.0

2.0 3.0 −1 (1000 / T) / K

4.0


43


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 44 of 47

Figure 6. (a) The potential barrier for the CH3 + CH4 à CH4 + CH3 reaction (R3) along the MEP of the RD PES (solid curve), and the semiclassical Eckart potential approximation (eq 11, dashed curve). (b) Rate constants for R3 calculated using TST (dashed red line) and RD SCTST (solid black curve), as well as scattering rate constants from Banks et al.44 (dashed blue curve). (c) RD SCTST and TST rate constants as compared to experimentally determined rate constants.71-72

44


Page 45 of 47

R3: CH3 + CH 4 0.05

CH4 + CH3

(a) V(ρ, δ) / a.u. MEP

−80.00

0.04

δ / rad

−80.04 0.03

CC path

T.S.

−80.08 −80.12

0.02

0.01 0.00

0.05

8.0

9.0

10.0 ρ / a.u.

11.0

12.0

(b) MEP SC approx CC parth

0.04

0.03

V / a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


T.S. 0.02

0.01

0.00 − 80

− 40 0 40 reaction coordinate / a.u.

80

Figure 7. (a) Contour plot of the fitted RD PES in hyperspherical coordinates from Banks et al.44 The red curve is the minimum-energy path (MEP), and the blue line represents a possible cornercutting (CC) path. (b) The potentials as a function of the reaction coordinate along each of these paths, as well as the semiclassical (SC) approximation to the potential (eq 11).

45



1.15

750

500

T/K

300

(a)

kRD SCTST-DT (T) / kRD SCTST (T)

1.10 1.05 1.00 0.95

R1 R2 R3

0.90 0.85 0.80 1.0

2.0

3.0

4.0

−1

(1000 / T) / K 0.003

(b)

R3: CH3 + CH4 T.S.

CH4 + CH3 SC barrier SC-DT barrier

0.002 V / a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 46 of 47

0.001

0.000 − 80

− 40

0 40 reaction coordinate / a.u.

80

Figure 8. (a) The ratio of the RD SCTST rate constants calculated with the deep tunneling corrections described in section 2.1.2 to those calculated without corrections for R1 (red curve), R2 (green curve), and R3 (blue curve). (b) The potential barrier for R3 used in deep tunneling RD SCTST calculations (blue curve), and the symmetric Eckart potential representing the potential barrier for the RD SCTST method without deep tunneling corrections (eq 11, red curve). In constructing these barriers, Ω{n} was set to |ħ ωF|.

46


Page 47 of 47

Table of Contents Image 0.020 0.015

V / a.u.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


R1: H + CH4

H 2 + CH3

T.S.

MEP

0.010 0.005 SC approx 0.000

− 40 − 20 0 20 40 reaction coordinate / a.u.

47


Semiclassical multi-phonon theory for atom-surface scattering: Application to the Cu(111) system.

Presolvated electron reactions with methyl acetoacetate: electron localization, proton-deuteron exchange, and H-atom abstraction.

Performance of PNOF6 for Hydrogen Abstraction Reactions.

Excited-state hydrogen atom abstraction initiates the photochemistry of β-2'-deoxycytidine.

Accounting for conformational flexibility and torsional anharmonicity in the H + CH3CH2OH hydrogen abstraction reactions: a multi-path variational transition state theory study.

Hydrogen abstraction from the hydrazine molecule by an oxygen atom.

Theoretical and kinetic study of the hydrogen atom abstraction reactions of esters with H(O.)2 radicals.

Direct and indirect hydrogen abstraction in Cl + alkene reactions.

Carbon-catalysed reductive hydrogen atom transfer reactions.

Kinetics of Hydrogen Radical Reactions with Toluene Including Chemical Activation Theory Employing System-Specific Quantum RRK Theory Calibrated by Variational Transition State Theory.

Crossing the dividing surface of transition state theory. II. Recrossing times for the atom-diatom interaction.

Reaction of an iron(IV) nitrido complex with cyclohexadienes: cycloaddition and hydrogen-atom abstraction.

Kinetics of hemoglobin and transition state theory.

Luminescent Ce(III) Complexes as Stoichiometric and Catalytic Photoreductants for Halogen Atom Abstraction Reactions.

Calculation of kinetic isotope effects for intramolecular hydrogen shift reactions using semiclassical instanton approach.

A hydrogen-atom abstraction model for the function of YZ in photosynthetic oxygen evolution.

Rate constant calculations of H-atom abstraction reactions from ethers by HȮ2 radicals.

Hydrogen-atom abstraction from a model amino acid: dependence on the attacking radical.

Application of the Zero-Order Reaction Rate Model and Transition State Theory to predict porous Ti6Al4V bending strength.

Dynamics of chlorine atom reactions with hydrocarbons: insights from imaging the radical product in crossed beams.

The Jastrow antisymmetric geminal power in Hilbert space: theory, benchmarking, and application to a novel transition state.

Properties and reactivities of nonheme iron(IV)-oxo versus iron(V)-oxo: long-range electron transfer versus hydrogen atom abstraction.

Mechanistic Studies of the Radical S-Adenosyl-L-methionine Enzyme 4-Demethylwyosine Synthase Reveal the Site of Hydrogen Atom Abstraction.

Thermal rate constants for the O(3P) + HBr and O(3P) + DBr reactions: transition-state theory and quantum mechanical calculations.